Solve differential equation with boundary conditions using substitution

Click For Summary
SUMMARY

The discussion focuses on solving the differential equation μ²(d²u/dx²) + ae^u = 0 with boundary conditions u(-L) = u(L) = u₀. Participants suggest using substitution and integration techniques to derive the solution. One proposed method involves multiplying by du/dx and integrating, leading to the equation m²u'² + 2a exp(u) = c. The final solution requires inverting x(u) to obtain u(x).

PREREQUISITES
  • Understanding of differential equations and boundary value problems
  • Familiarity with integration techniques in calculus
  • Knowledge of substitution methods in solving differential equations
  • Basic concepts of exponential functions and their derivatives
NEXT STEPS
  • Study the method of integrating factors for solving differential equations
  • Learn about boundary value problems and their applications in physics
  • Explore numerical methods for solving differential equations with boundary conditions
  • Investigate the use of software tools like MATLAB for solving complex differential equations
USEFUL FOR

Mathematicians, physicists, and engineering students who are working on differential equations and boundary value problems will benefit from this discussion.

cheesecake91
Messages
1
Reaction score
0
μ^{2}\frac{d^{2}u}{dx^{2}}+ae^{u}=0

Boundary conditions: u(-L)=u(L)=u_{0}

Solve by multiplying by \frac{du}{dx} and integrating in x

I know you have to use substitution, but I keep going in circles.
 
Physics news on Phys.org
cheesecake91 said:
μ^{2}\frac{d^{2}u}{dx^{2}}+ae^{u}=0

Boundary conditions: u(-L)=u(L)=u_{0}

Solve by multiplying by \frac{du}{dx} and integrating in x

I know you have to use substitution, but I keep going in circles.

The way that I would do it would be to rewrite the equation as
\mu^2 '' = -ae^\mu

Then you solve for a general solution which is
\mu(x) = 1 + x

Then you go on to get your general solution from the boundaries and such.

Another method, which would probably be easier and is what I think you were trying to do is to just integrate twice and get:
\frac{\mu^4}{12} = -ae^\mu + \mu*c_1 + c_2

That may be slightly off, but try integrating it yourself, and then use your boundary conditions to solve for the constants.
 
Last edited:
Hi !
I think that the two methods proposed by danielu13 are both incorrect.
Better follow the advice given in the wording : multiply by u'
m² u'' +a exp(u) = 0
m² 2 u''u' +2a exp(u) u' = 0 and integrate :
m² u'² +2a exp(u) = c
m u' = (+ or -)sqrt(c-2a exp(u))
Case +sqrt :
m (du/dx) = sqrt(c-2a exp(u))
dx = m du / sqrt(c-2a exp(u))
Then integrate, which gives x(u)
and invert it to obtain u(x).
 

Similar threads

  • · Replies 22 ·
Replies
22
Views
4K
  • · Replies 4 ·
Replies
4
Views
3K
  • · Replies 13 ·
Replies
13
Views
3K
  • · Replies 4 ·
Replies
4
Views
4K
  • · Replies 2 ·
Replies
2
Views
3K
  • · Replies 1 ·
Replies
1
Views
2K
  • · Replies 8 ·
Replies
8
Views
4K
  • · Replies 1 ·
Replies
1
Views
2K
Replies
1
Views
2K
  • · Replies 6 ·
Replies
6
Views
3K